The long term objective is to manipulate tumor-specific immunity in late tumor-bearing (LTB) individuals by changing the local environment of established tumors. It will be determined: (a) which types of T cells resist the adverse effects of a large tumor burden and retain the capacity to destroy cancers, (b) how a large tumor burden may affect the localization and maturation of tumor-specific T cells, (c) whether TGF-beta produced by tumor or stromal cells contributes to progressive tumor growth and (d) whether effective T cells can be rescued from LTB. It will also be determined (a) which stromal components influence the transplantability of solid tumors and, (b) whether genetic or other alterations of stromal cells can lead to destruction of advanced cancers. Syngeneic tumors will be used that retain antigens defined by CD8+ or CD4+ T cell clones. T cells from tumor-free mice, tumor-specific CD4+ T cell clones differing in secreted cytokines, and short-term cultured T cells from tumor-immune mice will be used to rescue CD8+ T cells from LTB. Limiting dilution analysis of tumor-infiltrating lymphocytes will determine whether tumor-specific T cells in LTB accumulate but fail to mature in the tumor. Also, it will be determined which stromal cells in the tumor produce TGF-beta, which cells activate TGF-beta and whether TGF-beta prevents the maturation of tumor-infiltrating T cells. In complementary studies, we will determine whether expression of a TGF- beta anti-sense construct by TGF-beta secreting tumors causes the tumor to regress. Thus, studies of the stromal environment of established tumors may allow the development of new approaches for manipulating effective tumor immunity.